READ ME FILE


This folder contains the zTree programs, experimental protocol instruction, raw data files, and cleaning and analysis .do files for the paper titled “Fairness Has Less Impact When Agents Are Less Informed” by Jennie Huang, Judd B. Kessler, and Muriel Niederle.


***The zTree programs are contained in the folder: zTree_game_program
Original experimental protocol instructions for the game can be found in this folder in PDF format (filename: Fairness_Experimental_Instructions.pdf). Additionally, experimental instructions can also be found in Appendix C of the paper. 
Only the first two parts of the experiment (files ending in _P1 and _P2) were relevant for the analysis done in this paper. 


***The raw data files (outputs from the zTree games) are contained in the folder: Raw_data
This folder contains 20 .xlsx files corresponding to the 20 experimental sessions, 8 of which were run at Stanford and 12 at Wharton. The file names corresponded to the timestamps of the session as YYMMDD_HHMM.xlsx 


The following sessions were run at Stanford: 120509_1658, 120509_1855, 120514_1712, 120516_1646, 120522_1938, 120524_1736, 120529_1820, 120604_1909


The remaining sessions were run at Wharton: 120614_1120, 120614_1248, 120614_1356, 120615_1109, 120615_1234, 120615_1419, 120618_1149, 120618_1350, 120619_1130, 120619_1343, 120620_1136, 120620_1345


Each session corresponds to 1 treatment group. Below we outline which treatment/version of the game was played in each session file.


Treatment 1 = Neither Knows (NK)
120509_1658, 120514_1712, 120614_1248, 120615_1109, 120619_1343


Treatment 2 = Seller Knows (SK)
120516_1646, 120529_1820, 120615_1234, 120618_1149, 120620_1345


Treatment 3 = Buyer Knows (BK)
120509_1855, 120524_1736, 120614_1356, 120618_1350, 120619_1130


Treatment 4 = Complete Information (CI)
120509_1855, 120524_1736, 120614_1356, 120618_1350, 120619_1130




***The .do files to clean and analyze the data are in the folder: Stata_do_files
In total there are 2 cleaning files (files 00_ and 01_) and 2 analysis files (files 02_ and 03_). 


*File 00_cleaning_fairness_game_part1.do is used to import the .xlsx raw data files and clean the outcomes from the main game (part 1) used in this paper. A cleaned data file named fairness_part1 in .dta format is produced at the end of this cleaning file. 

Key variables in this fairness_part1.dta file are listed below. 


sessionfile — "Session file number"
part — "Experiment part; 4 parts total"
period — "Period; 30 periods total"
subjectnum — "zTree subject number"
group — "zTree group identifies pairs"
role — "Role of subject; Buyer==1, Seller==2"
value — "Value of Buyer; Randomly selected 70 or 90"
cost — "Cost of Seller; Randomly selected 10 or 30"
othercost — "Cost of seller when subject is a buyer" 
othervalue — "Value of buyer when subject is a seller"
offer — "Buyer's offer"
accept — "Seller accept or reject; accept==1, otherwise 0"
earnings — "Buyer and seller earnings"
otherearnings — "Partner's earnings"
session  — "Session number" (Numeric 1-20, assigned based on chronological order by the session time stamp)
treatment — "Treament: 1==NK; 2==SK; 3==BK; 4==CI" (Numeric 1-4, corresponding to the 4 treatments, assigned based on the session files, see notes above)
subj_id — "Subject ID" (Generated by combining session*100 + subjectnum)
Buyer — "Buyer==1, otherwise 0" (Dummy flag for buyer where role == 1)


*File 01_cleaning_fairness_game_part2.do is used to import the .xlsx raw data files and clean outcomes from the belief elicitation portion of the game (part 2) used in this paper. A cleaned data file named fairness_part2 in .dta format is produced at the end of this cleaning file. Key variables in this fairness_part2.dta file used in our analysis are listed below. (Note: While there are other variables, they are not used in the paper.)


sessionfile — "Session file number"
part — "Experiment part; 4 parts total"
period — "Period; 10 periods total in part 2"
subjectnum — "zTree subject number"
group — "zTree group identifies pairs"
role — "Role of subject; Buyer==1, Seller==2"
guessperiod — "Period from Part 1 current player is guessing"
oldvalue — "Value of Buyer from Part 1; Randomly selected 70 or 90"
oldaccept — "Seller accepted offer from Part 1; oldaccepted==1, otherwise 0"
oldoffer — "Offer from Part 1"
oldcost — "Cost of Seller from Part 1; Randomly selected 10 or 30"
probBDM — "Buyer's guess of probability seller accepted offer; Seller's guess of probability buyer has value==70"



*File 02_analysis_paper_figures.do and 03_analysis_paper_tables.do read in fairness_part1.dta and fairness_part2.dta and produced the analysis shown in this paper. The first file is used to produce all figures and the second file is used to produce all tables and regressions shown in the main text and appendix.